Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Fair AI-STA for Legacy Wi-Fi: Enhancing Sensing and Power Management with Deep Q-Learning

Peini Yi, Wenchi Cheng, Zhanyu Ju, Jingqing Wang, Jinzhe Pan, Yuehui Ouyang, Wei Zhang

TL;DR

The paper tackles fairness and QoS in legacy Wi-Fi networks by introducing an AI-STA that uses Deep Q-Learning to adapt its carrier sensing threshold, receiver sensitivity threshold, and transmit power. It formulates an MDP with a discretized three-dimensional action space and a reward structure that encodes fairness via Jain's coefficient and QoS constraints, then trains a DQN with experience replay and target networks. A NAV-based, STA-focused independent fairness metric is proposed to evaluate channel-time sharing, alongside standard QoS metrics (Throughput, Latency, Jitter, Packet Loss) assessed through discrete-event ns-3 simulations. Results indicate that the DQN-based AI-STA can outperform DSC and baseline 802.11ax in fairness and throughput, though delay and jitter can increase due to limited MAC-parameter optimization, underscoring the potential benefits of incorporating additional MAC controls. The work demonstrates a robust framework for AI-driven sensitivity and power management in legacy Wi-Fi environments and highlights avenues for further QoS enhancement through broader MAC parameter optimization.

Abstract

With the increasing complexity of Wi-Fi networks and the iterative evolution of 802.11 protocols, the Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) protocol faces significant challenges in achieving fair channel access and efficient resource allocation between legacy and modern Wi-Fi devices. To address these challenges, we propose an AI-driven Station (AI-STA) equipped with a Deep Q-Learning (DQN) module that dynamically adjusts its receive sensitivity threshold and transmit power. The AI-STA algorithm aims to maximize fairness in resource allocation while ensuring diverse Quality of Service (QoS) requirements are met. The performance of the AI-STA is evaluated through discrete event simulations in a Wi-Fi network, demonstrating that it outperforms traditional stations in fairness and QoS metrics. Although the AI-STA does not exhibit exceptionally superior performance, it holds significant potential for meeting QoS and fairness requirements with the inclusion of additional MAC parameters. The proposed AI-driven Sensitivity and Power algorithm offers a robust framework for optimizing sensitivity and power control in AI-STA devices within legacy Wi-Fi networks.

Fair AI-STA for Legacy Wi-Fi: Enhancing Sensing and Power Management with Deep Q-Learning

TL;DR

The paper tackles fairness and QoS in legacy Wi-Fi networks by introducing an AI-STA that uses Deep Q-Learning to adapt its carrier sensing threshold, receiver sensitivity threshold, and transmit power. It formulates an MDP with a discretized three-dimensional action space and a reward structure that encodes fairness via Jain's coefficient and QoS constraints, then trains a DQN with experience replay and target networks. A NAV-based, STA-focused independent fairness metric is proposed to evaluate channel-time sharing, alongside standard QoS metrics (Throughput, Latency, Jitter, Packet Loss) assessed through discrete-event ns-3 simulations. Results indicate that the DQN-based AI-STA can outperform DSC and baseline 802.11ax in fairness and throughput, though delay and jitter can increase due to limited MAC-parameter optimization, underscoring the potential benefits of incorporating additional MAC controls. The work demonstrates a robust framework for AI-driven sensitivity and power management in legacy Wi-Fi environments and highlights avenues for further QoS enhancement through broader MAC parameter optimization.

Abstract

With the increasing complexity of Wi-Fi networks and the iterative evolution of 802.11 protocols, the Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) protocol faces significant challenges in achieving fair channel access and efficient resource allocation between legacy and modern Wi-Fi devices. To address these challenges, we propose an AI-driven Station (AI-STA) equipped with a Deep Q-Learning (DQN) module that dynamically adjusts its receive sensitivity threshold and transmit power. The AI-STA algorithm aims to maximize fairness in resource allocation while ensuring diverse Quality of Service (QoS) requirements are met. The performance of the AI-STA is evaluated through discrete event simulations in a Wi-Fi network, demonstrating that it outperforms traditional stations in fairness and QoS metrics. Although the AI-STA does not exhibit exceptionally superior performance, it holds significant potential for meeting QoS and fairness requirements with the inclusion of additional MAC parameters. The proposed AI-driven Sensitivity and Power algorithm offers a robust framework for optimizing sensitivity and power control in AI-STA devices within legacy Wi-Fi networks.

Paper Structure

This paper contains 29 sections, 27 equations, 7 figures, 1 table, 2 algorithms.

Table of Contents

  1. Introduction
  2. System Model
  3. Network Model
  4. Channel Model
  5. Link Model of AI-STA Model
  6. Uplink
  7. Downlink
  8. Optimization Target
  9. Independent Fairness Metric
  10. QoS Performance
  11. Throughput
  12. Latency
  13. Jitter
  14. Packet Loss Rate
  15. Problem Formulation
  16. ...and 14 more sections

Figures (7)

  • Figure 1: The single AI-driven STA in a legacy Wi-Fi network scenario
  • Figure 2: The AI-STA Listening to RTS/CTS Packets and Updating STA-Duration Table
  • Figure 3: The DQN with varying of time
  • Figure 4: CST comparison of DQN-based, DSC, and baseline algorithm varying of time.
  • Figure 5: Throughput comparison of DQN-based, DSC, and baseline algorithm varying of time.
  • ...and 2 more figures